Difference-Based Relational Learning for Zero-Shot Object-Goal Visual Navigation With Direct Sim-to-Real Transfer
On the map: Deep reinforcement learning methods → RL for robot navigation and UAV control
Abstract (arXiv)
End-to-end deep reinforcement learning (DRL) for zero-shot object-goal visual navigation remains challenged by the sim-to-real gap, particularly variations in object appearance and restricted camera field-of-view (FoV). This letter proposes a Temporal Difference-Relational Network (T-DRN) for robust zero-shot sim-to-real transfer. T-DRN combines a Siamese difference-based feature extractor, which computes relational difference between the target and observed objects to produce domain-independent representations, with a dual-frame temporal buffer that preserves short-term object continuity under narrow FoV. Extensive experiments in AI2-THOR demonstrate that T-DRN improves zero-shot generalization in terms of success rates over strong baselines. Furthermore, T-DRN is systematically validated on a physical wheeled robot, demonstrating robust performance under real sensing and actuation constraints and supporting the feasibility of direct sim-to-real transfer.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods. Builds on 17 of its 26 references that are in the map. First posted on arXiv 17 Jul 2026.
Builds on
17 of 26 references.
Closest ancestors · 1
Same direction, newest first
- 2024TDANet: Target-Directed Attention Network For Object-Goal Visual Navigation With Zero-Shot Ability12 Apr
Other references · 16
- 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
- 2022YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectorscited by 982
- 2017AI2-THOR: An Interactive 3D Environment for Visual AIcited by 869
- 2018On Evaluation of Embodied Navigation Agentscited by 547
- 2020Object Goal Navigation using Goal-Oriented Semantic Explorationcited by 429
- 2016Target-driven Visual Navigation in Indoor Scenes using Deep Reinforcement Learningcited by 366
- 2019Scene Memory Transformer for Embodied Agents in Long-Horizon Taskscited by 107
- 2018Learning to Learn How to Learn: Self-Adaptive Visual Navigation Using Meta-Learningcited by 93
- 2020Learning Object Relation Graph and Tentative Policy for Visual Navigationcited by 64
- 2024VoroNav: Voronoi-based Zero-shot Object Navigation with Large Language Modelcited by 57
- 2022Zero-shot object goal visual navigationcited by 31
- 2022Object Memory Transformer for Object Goal Navigationcited by 20
4 more
- 2022BioSLAM: A Bio-inspired Lifelong Memory System for General Place Recognitioncited by 13
- 2024Aligning Knowledge Graph with Visual Perception for Object-goal Navigationcited by 9
- 2025osmAG-LLM: Zero-Shot Open-Vocabulary Object Navigation via Semantic Maps and Large Language Models Reasoningcited by 3
- 2021Image-Based Trajectory Tracking through Unknown Environments without Absolute Positioningcited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2025VLD: Visual Language Goal Distance for Reinforcement Learning Navigation
- 2026120 Minutes and a Laptop: Minimalist Image-goal Navigation via Unsupervised Exploration and Offline RL
- 2026T2Nav Algebraic Topology Aware Temporal Graph Memory and Loop Detection for ZeroShot Visual Navigation ↗
- 2025NavigateDiff: Visual Predictors are Zero-Shot Navigation Assistants
- 2026Sim-to-Real Transfer via a Style-Identified Cycle Consistent Generative Adversarial Network: Zero-Shot Deployment on Robotic Manipulators through Visual Domain Adaptation
- 2025Zero-Shot Visual Generalization in Robot Manipulation
- 2026C2Nav: Compare Before You Commit for Zero-Shot Vision-and-Language Navigation
- 2024LMD-PGN: Cross-Modal Knowledge Distillation from First-Person-View Images to Third-Person-View BEV Maps for Universal Point Goal Navigation